Problem with using periodic boundary conditions
 

Hi,
I am simulating one-sixth of an unbaffled mixer with a six-blade impeller. This part (one-sixth) consists of just one blade. I am using MRF for modeling of the rotation and periodic boundary conditions for the side walls.
The problem is by assigning periodic BC for these two walls, the whole vessel rotates with the same velocity and the profiles of tangential velocities in different parts of the vessel are the same, which deviate from the experimental data.

Was wondering if someone can give me some pieces of advice to handle this problem.

Thanks,